Automatic quantitation of vacuolar lesions in the brain of mice infected with transmissible spongiform encephalopathies

The spongiform change induced in the brain tissue is one of the major features investigated in the neuropathology in natural and experimental transmissible spongiform encephalopathies (TSEs). In this context, a reproducible quantitation of the magnitude of these vacuolar lesions is described using image analysis techniques. To be exploited successfully, this image analysis must be able to distinguish the vacuolar lesions from vascular elements. The present study describes the different calibration stages of image analysis using hematoxylin-eosin stained slices of brain from mice infected with TSEs. In order to select automatically a maximum of vacuoles and to eliminate a maximum of vascular elements, relevant criteria based on the threshold values of the optical density, shape factor and surface of detected objects were determined. Compared to visual scoring, this method has the advantages of enhanced precision of the measure, reproducibility and moreover, the collection of numerical data for more detailed statistical analysis. In addition, an original scale change function is proposed allowing a comparative analysis with values from the visual scoring method. The method of automatic recognition and quantitation of vacuolar lesions described in this paper represent a useful tool for large-scale analysis of spongiform lesions induced by different TSE isolates transmitted to mice. (C) 2004 Elsevier B.V. All rights reserved.